I would first check the wheel bearing and condition of each track-rod-end and ball joints. If either has 'play' then sort the 'play' before assuming its a tracking fault.

You need a length of timber approx 50x30mm that is rigid. Cut the length to fit between the front wheels. Two bullnose clips. Two lengths of thin rigid steel or plastic, approx 300 x 15mm x 2mm which are used as 'pointers'.
Place the timber horizontal and secure the 'pointers' towards the ends of the timber so they point vertical ....... 90 degrees.
Now the method:
Tracking is measured across the inside face of the front wheels. Correct position is onto the wheel rim at centre height.The car must have its wheels on the floor. Consequently the use of an inspection pit makes the job easy.
Set the timber onto the floor, across the car, between the baulkhead and wheels. Position the pointers to touch each rim of the front wheels - centre height of rim. Now very carfully remove the timber and reposition forward of the front wheels. Dont move the pointers ... takes a few attempts! Forward of the wheels and set in position it is possible using a small ruler/judgement to measure the difference ... either its tracking in/out or same. You can adjust the track-rod-ends and set the track to manufacturers specifications. It takes a few attempts but the accuracy is very acceptable. NB ... adjust the wheel which is showing ware. The other wheels is probably OK.
If each track-rod is adjusted then by looking at the length of thread visible on each track-rod-end it is possible to judge where any adjustment is needed.
The method is not as reliable as when a garage tracks wheels with view to aligning with the rear wheels. If you have knocked one wheel then my method proves accurate. However, I have tracked many wheels which proved to be within 1 degree of accuracy. Good enought and at no cost.